Chronic Poverty: A Review of Current Quantitative Evidence

Abstract

Many aspects of living conditions can fluctuate significantly from one period of time to another, as for example the income levels or nutritional status of small scale agricultural producers subject to the effects of climatic variability or price changes. Therefore in discussing poverty, it is highly desirable to adopt a dynamic perspective, and draw a distinction between transient and chronic poverty. For those experiencing transient poverty their poverty is temporary; such individuals may experience movements into and out of poverty but are not persistently poor. The chronically poor by contrast experience persistent poverty over a reasonably long period of time. As in the analogous case of short-term and long-term unemployment, these two types of poverty may be quite distinct and call for different policy responses.

This paper discusses available quantitative evidence in relation to chronic poverty. After an introduction, section 2 discusses conceptual issues in defining and measuring chronic poverty. To identify chronic poverty will generally require either comparable information on individuals at two or more points in time (longitudinal or panel data) or information that though only collected once, provides information on dynamics (e.g. retrospective information). Much of the literature has focused on panel data sets and monetary measures of living conditions; this section discusses the two approaches that have been used in such studies to the chronically poor (the spells and components approaches). It also highlights the importance of measurement error in making this distinction and the need to identify chronic and transient aspects of deprivation in non-monetary dimensions where this distinction is relevant. Available evidence from panel data sets on the extent of chronic poverty and the characteristics of the chronically poor is reviewed in section 3.

In almost all instances a minority of the poor at a given point in time are identified as being chronically poor, though often this is still a substantial minority. The chronically poor are typically associated as having distinct characteristics which may explain their persistent poverty, such as low levels of human capital and productive assets, being in households with high dependency ratios and working in low return activities. By contrast the transient poor are those that have difficulty in insuring themselves against the consequences of shocks such as adverse price changes. In many cases though panel data are not available, and even where they are, they inevitably suffer from various limitations - in particular the difficulty of measuring changes accurately at the individual level. Section 4 therefore considers various ways in which it may be possible to consider chronic poverty in the absence of panel data. Section 5 concludes, including by summarizing important issues for future research.

Keywords: poverty dynamics, international comparisons, panel data